The quality of life and survival for cancer patients are considerably threatened by liver metastases. While surgical resection is often not possible and systemic chemotherapy is largely ineffective, promising interventional therapeutic techniques are available. The focus of this project is to demonstrate the effectiveness of real-time time-resolved 3D MR guidance for one such technique: transcatheter arterial chemoembolization (TACE). Successful completion of the project will not only allow immediate improvement of TACE procedure, but should prove to be a capable foundation for other drug delivery techniques where fusing information of local vasculature and anatomy is beneficial. Current guidance for TACE from X-ray Digital Subtraction Angiography( DSA) provides thick 2D projection images with little information on tumor location while increasing the already high dose of ionizing radiation received by liver tumor patients. While MR imaging has provided 3D capabilities for some time, further improvements in acquisition speed, reconstruction processing, and visualization are necessary to apply real-time 3D MRI to new problems. Challenges are derived from the extraordinary increase in data created by 4D imaging. But opportunities are derived from the rapid development of parallel imaging in MRI, new paradigms for using MR's multiple contrast mechanisms, and the continued dramatic improvement in computing power. This project aims to develop 4D MRI, three spatial dimensions plus time, to better visualize the structure and significance of the vasculature adjacent to tumors, and integrate it with the other interventional imaging methods necessary for TACE. Specifically, the first two specific aims of this project concentrate on reconstruction and visualization: 1) Improve MR imaging speed and coverage using faster, more efficient non-Cartesian techniques and integrating these into a real-time reconstruction platform and 2) visualizing of 4D data using a computer cluster. The third aim integrates these developments with other capabilities necessary for MR guidance including catheter visualization and 3D T2-weighted imaging for lesion conspicuity. The last aim verifies these capabilities using the same methods used to validate clinical TACE. Successful completion will provide the necessary information to support a later human trial and a platform for guiding TACE as well as drug delivery in general. Outcomes for the first two aims will also shorten acquisition time, improve image quality, and simplify interpretation of time-resolved diagnostic MR angiography.